For rule-based investors
You run your own money with rules and you have been burned by a backtest that looked great and did not work. Stoxa runs every rule against what was public on each date, nothing more, and now takes you from a screen to a ranking, a backtest, a portfolio simulation and a strategy tracked live every night.
Most datasets show today's version of the past. Stoxa stores every fact with the date it became public, so any question can be asked as of any day.
A quarter ends in March, but the 10-Q lands weeks later, and figures are sometimes restated after that. A screen or backtest that uses the period-end date, or the latest restated value, trades on information nobody had yet. The results look better than anything you could have done.
Every fundamental carries its period end, its filing date and its accession number. A query "as of D" sees only filings with a filing date on or before D, and takes the latest revision known at that point. Restatements are stored as new rows, never written over the original.
A backtest runs your screen on each rebalance date against the data as it stood that day, then measures forward returns. Macro inputs work the same way: the risk-free rate and regime columns come from the vintage that was published at the time.
filed <= as_of, latest revision wins.
Screening, ranking, backtesting, simulation and macro share the same as-of date, so every part of an analysis sees the same world. Each step feeds the next: a screen becomes a universe, a ranking orders it, a backtest tests it and the simulator runs it as a portfolio.
Filter US-listed equities on fundamentals, multiples, technicals and factor ranks, as of any past date, over SEC EDGAR data.
Order stocks with a weighted tree of factors and composites, each ranked within the universe, the sector or the industry. Eight built-in systems to copy and edit.
Run a screen or a ranking across rebalance dates using only data public on each date, and measure what it would have earned.
Run a strategy as a portfolio: positions are bought once and held until a rule sells them, with cash, dividends and costs on the trades actually made.
Eight house strategies run by the same simulator, from a multi-factor combination to the Magic Formula. Their rules are readable in full.
Each value traces back to its filing, accession number and fetch date. Raw source responses are kept, so everything can be re-derived.
Each strategy is a fixed rule set run by the portfolio simulator and tracked every night from its launch date. This table shows only the record since launch. The backtest before launch is on each strategy's page, labelled as a backtest, because the rules were chosen knowing that period.
| Strategy | Holdings | Since launchvs SPY |
|---|---|---|
| Core Combination | 25 | Live since2 Nov 2026 |
| Quality at a Reasonable Price | 20 | Live since2 Nov 2026 |
| Magic Formula | 30 | Live since2 Nov 2026 |
| Piotroski-style Value | 20 | Live since2 Nov 2026 |
| Dividend Growth | 20 | Live since2 Nov 2026 |
| Low-Volatility Quality | 30 | Live since2 Nov 2026 |
| Momentum with a Quality Floor | 20 | Live since2 Nov 2026 |
| Market-Implied Growth Gap | 20 | Live since2 Nov 2026 |
A simulation is only useful if you know what it leaves out. Every rule is in the simulation guide.
Macro data gets revised, sometimes heavily. Stoxa keeps every vintage, so the CPI print you see as of a date is the one that was out that day. The same store sets the risk-free rate in your WACC and adds regime columns to screens and backtests.
The recession signal is the real-time Sahm rule, not NBER dating. NBER announces recessions up to a year after the fact, so using its dates in a backtest would leak the future. They are only used to shade charts.
Portfolio123 is the other tool that gets point-in-time data right, and the one these features are measured against. It has a longer history, global data and a formula language; Stoxa has a valuation engine, open strategy rules and one flat price.
| Feature | Stoxa | Portfolio123 |
|---|---|---|
| Point-in-time fundamentals | Yes | Yes |
| Restatements kept, not overwritten | Yes | Partial |
| Ranking systems with a bucket test | Yes | Yes |
| Portfolio simulation with stops and hedging | Yes | Yes |
| Ready strategies with rules shown in full | Yes, eight | Designer models, rules hidden |
| Custom formula language | No | Yes |
| Fundamentals history | SEC XBRL era (2009+) | 1999+, depth by plan |
| Valuation engine included | Yes | No |
| API included | Yes, no credits | Credits capped by plan |
| Price | $29/mo | ~$35–278/mo by plan |
As published on portfolio123.com on 1 November 2026 (non-professional plans, one region); check it for current terms.
$29 a month or $290 a year. The full workbench, the API and the Python SDK. No data add-ons. Cancel any time.
$29/month
Sign in and subscribe$290/year
Sign in and subscribeUS-listed companies filing with the SEC (no non-US coverage yet). Questions: hello@stoxa.io.